Stochastic Modeling and Inference with Multi-type Branching Processes

Date Thursday November 17, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium
Speaker Jason Xu, Ph.D., Postdoctoral Fellow, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Markov branching processes are a class of continuous-time Markov chains (CTMCs) with ubiquitous scientific modeling applications. Multi-type processes are necessary to model phenomena such as competition, predation, or infection, but often feature large or uncountable state spaces, rendering standard CTMC techniques impractical. We present recent methodology that enables likelihood computations and EM algorithms in these settings. We examine the performance of these techniques applied to data from molecular epidemiology and hematopoiesis studies, and briefly explore alternatives when such methods are limited, including moment-based estimators and compressed sensing techniques that scale to large systems and datasets.
Flyer Jason_Xu_20161117.pdf